Scholarlink

<p>Reliably estimating streamflows is paramount for informed decision-making in water resources management. In this sense, properly accounting for uncertainty in rating curves is a necessary step for characterizing the streamflow regime at a river cross section, particularly when complex morphological changes are observed over time. In this paper, we investigate the suitability of the nonstationary stage-period-discharge (SPD) approach, a Bayesian framework built upon prior hydraulic knowledge on the flow regimes and designed for dynamically assessing the evolution of some of the rating curve parameters, for modeling the stage-discharge relationships of a highly unstable river reach in Brazil. We tested SPD under two structural error models, which differ in the treatment of low flow uncertainty, and across 14 periods of validity. We also assessed the model predictive skills by transferring information on time-varying parameters for future (unobserved) rating changes. Our results indicated that, even under a complex rating change dynamic, SPD could properly capture the channel bed evolution and entailed relatively accurate rating curves in all periods. However, the rating changes strongly affected the inference of the rating curve activation stages, leading to bias on the higher flows that could only be addressed with additional high gauges. On the other hand, the model was less effective during prediction, with fairly unprecise and biased streamflow estimates even for short lead times, which limits its application in practical hydrology.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Assessing the stage-period-discharge framework for modeling highly unstable rating curves: a case study at the UHE Peti Carrapato streamflow Gauging Station, Brazil

  • 期刊论文

  • 发表日期：

    2025年1月19日

查看全文
  • Iago Eleutério,
  • Francisco Silva,
  • Veber Costa

摘要

Reliably estimating streamflows is paramount for informed decision-making in water resources management. In this sense, properly accounting for uncertainty in rating curves is a necessary step for characterizing the streamflow regime at a river cross section, particularly when complex morphological changes are observed over time. In this paper, we investigate the suitability of the nonstationary stage-period-discharge (SPD) approach, a Bayesian framework built upon prior hydraulic knowledge on the flow regimes and designed for dynamically assessing the evolution of some of the rating curve parameters, for modeling the stage-discharge relationships of a highly unstable river reach in Brazil. We tested SPD under two structural error models, which differ in the treatment of low flow uncertainty, and across 14 periods of validity. We also assessed the model predictive skills by transferring information on time-varying parameters for future (unobserved) rating changes. Our results indicated that, even under a complex rating change dynamic, SPD could properly capture the channel bed evolution and entailed relatively accurate rating curves in all periods. However, the rating changes strongly affected the inference of the rating curve activation stages, leading to bias on the higher flows that could only be addressed with additional high gauges. On the other hand, the model was less effective during prediction, with fairly unprecise and biased streamflow estimates even for short lead times, which limits its application in practical hydrology.